Separate Timing Mechanisms for Duration and Rate Perception: Evidence and Implications

Abstract

Temporal perception encompasses several timing properties, including duration and temporal rate. If temporal perception were a unified mechanism, one might predict that differences in perception would apply across properties; thus, a stimulus that seems to persist for a longer duration might also seem to flicker more slowly. Moreover, effects that seem to be largely unisensory for duration might also be expected to be sensory specific for rate. However, a growing body of evidence suggests that the mechanisms that underlie perception of temporal rate work quite differently from those that underlie duration perception, which has important implications for models of time perception.In a series of experiments, Johnston, Arnold, and Nishida (2006) showed that they could manipulate perceived duration of a visual stimulus using an adaptation paradigm. They found that their adaptation was specific to particular locations and that the effect on duration could be dissociated from perception of the rate of flicker of their stimuli. They argued that the neural mechanisms involved in duration and temporal frequency must be different.This contention is further supported from results using a chronostasis paradigm. With chronostasis, under certain conditions, a perceptual expansion of time occurs and the duration of certain stimuli are overestimated. For instance, when streams of stimuli are presented, an “oddball” stimulus that differs from a set of identical standards (e.g., altered in shape, motion, color, or size) is perceived to last longer than the standard stimuli (Tse, Intriligator, Rivest, & Cavanaugh, 2004). While these oddball effects are robust for duration, judgments of flicker rate are unchanged by oddball status (Pariyadeth & Eagleman, 2007). Although the cause of chronostasis is still under debate, the difference between the duration and rate experimental results again suggests a separability between the types of tasks.Newer evidence for this separability of rate and duration perceptual tasks comes from examining adaptation studies that looked for transfer from one sensory modality to another. Exposure to auditory or visual adaptors of a particular duration leads to an aftereffect such that short adaptors lead to a perceived expansion of duration, whereas long adaptors lead to compression. This aftereffect is tuned such that the expansion and contraction occur only for durations in a limited range. This duration aftereffect is not due to temporal frequency adaptation and is sensory-specific; an adaptor in a particular modality has no effect on test stimuli in the other modality (Heron et al., 2012). Studies of temporal rate adaptation have also demonstrated a negative aftereffect such that adaptation stimuli pulsing quickly lead to subsequent stimuli seeming to pulse more slowly, and slow adaptors lead to later stimuli seeming to pulse more quickly (Levitan, Ban, Stiles, & Shimojo, 2012). This aftereffect shows tuning similar to that found for duration (Levitan et al., 2013), but unlike the duration results, the effect of adaptation transfers across modalities – auditory adaptors change the perception of visual test stimuli, and visual adaptors change the perception of auditory test stimuli. Taken together, these findings suggest that mechanisms underlying duration perception are modality-specific, while mechanisms involved in rate perception apply across modalities.At the neural level, a pair of recent electrophysiological studies further support the notion for a distinction between duration and frequency mechanisms. For duration discrimination, the neural circuitry involved in auditory processing is distinct from the circuitry involved in somatosensory processing (Butler et al., 2011). This result argues for unisensory representation of duration. However, the same research group used a similar paradigm to probe rate perception, and found significant multisensory coupling, suggesting that for rate, the modalities interact early on, even in a task that does not require directed attention (Butler et al., 2012). These neural results are consistent with the psychophysical evidence for sensory-specific duration perception but cross- modal mechanisms for rate.These results provide some constraints upon possible models of temporal perception. Early models of temporal perception posited a clock-like mechanism, in which a pacemaker generates “ticks” that are counted by an accumulator (e.g., Treisman, 1963; see also Buhusi & Meck, 2005, for further discussion of such models). Such models cannot provide a reasonable account for the discrepancies between duration and rate; if temporal adaptation changes a multisensory tick rate, then this should influence the final assessment of duration. Non-clock models of temporal perception that consider the roles of sensory modalities in encoding time (e.g., van Wassenhove, 2009) must account for this divergence between duration and rate.